1. Field of the Invention
The present invention relates to a method and a device for automatically detecting a data transmission unit with a duplicated network address, and more particularly to a method and a device for automatically detecting a data transmission unit with a duplicated address on a network under control of link state routing.
2. Description of the Related Art
Conventionally, this type of address duplication detection system has been used for detecting the duplicated address of a data transmission unit on a local area network (LAN). An earlier patent disclosure dealing with this is found in Japanese Patent Publication (unexamined) No. A-3-212038.
FIG. 8 shows an example of the configuration of a LAN system using a conventional address duplication detection system.
FIG. 9 shows the format of a terminal initial packet. This terminal initial packet is sent either by a LAN terminal 28 in the LAN system shown in FIG. 8 when it starts sending packets or by a terminal adapter 29 when it detects that a line terminal 30 connected to it starts sending packets. The terminal initial packet is composed of the destination Media Access Control (MAC) address, the MAC address of the terminal, and the subscriber""s address.
FIG. 10 shows an address table stored in a network monitoring unit 26. When the terminal initial packet is received, the system searches the address table for a match. If the subscriber""s address in the terminal initial packet is not found in the table, the pair of the subscriber""s address and the MAC address is added to the table.
Now, assume that a LAN terminal 35 is added to the LAN system, shown in FIG. 8, with its subscriber""s address mistakenly set to the address of a line terminal 34, one of the existing terminals. The following describes how the conventional address duplication detection system works in this situation. In the following description, the subscriber""s address of the LAN terminal 35 is AAAA, its MAC address is 111111111111, the subscriber""s address of the line terminal 34 is AAAA, and the MAC address of the terminal adapter 29 is 222222222222.
When the line terminal 34 starts operation, the modem interface signal is activated and sent to the terminal adapter 29. Upon receiving this signal, the terminal adapter 29 sends the terminal initial packet in the format shown in FIG. 9 to the network monitoring unit 26 via the transmission line. In this case, the terminal initial packet contains the MAC address of 222222222222 and the subscriber""s address of AAAA.
Upon receiving the terminal initial packet, the network monitoring unit 26 checks if the subscriber""s address is already in the address table and if the MAC address stored in the address table matches the MAC address of the received terminal initial packet. In this case, the network monitoring unit 26 finds that the both are satisfied and then ends duplication checking. If the subscriber""s address is not stored in the address table, the network monitoring unit 26 adds the pair of the subscriber""s address and the MAC address to the address table.
On the other hand, when the LAN terminal 35 starts operation, it sends a terminal initial packet containing the MAC address of 111111111111 and the subscriber""s address of AAAA to the network monitoring unit 26. Upon receiving this terminal initial packet, the network monitoring unit 26 finds that the subscriber""s address is already in the address table but that the MAC address stored in the address table does not match the MAC address contained in the received terminal initial packet. That is, the network monitoring unit 26 detects a subscriber""s address duplication.
The automatic address-duplication detection device disclosed in Japanese Patent Publication (unexamined) No. A-3-267845 is used on a ring LAN and in a data transmission unit connected to the ring LAN. To detect address duplications, this device uses two types of node address for each data transmission unit: a node address which is unique and is usually set in the unit and a test node address which is specifically prepared for testing and which does not duplicate with the node address. First, the device in the data transmission unit sends onto the ring LAN a line loopback frame containing the unique node addresses as the destination node address and as the sending node address. If the loopback frame is not sent back to the sending data transmission unit, then the device sends onto the ring LAN a line loopback frame containing the test node address as the destination node address and as the sending node address. If the frame is sent back to the sending transmission unit, then the device finds that a transmission unit with the node address of the sending transmission unit is on the network.
The systems and device according to the prior art have the following problems. The first problem is that, when a commercially-available router not compatible with the system disclosed in Japanese Patent Publication (unexamined) No. A-3-212038 is used to connect LANs, a remote LANs also requires a network monitoring unit in order to detect address duplications. The second problem with this system is that, if a network monitoring unit fails due to a power failure or an operation error, address duplications cannot be detected.
The problem with Japanese Patent Publication No. A-3-267845 is that the device detects a duplication only in the termination (or sending) unit but not in intermediate systems (routers).
It is an object of this invention to provide a method for automatically detecting a network address duplication in network addresses assigned to data transmission units (intermediate systems) which are installed not only in one LAN but also on a point-to-point line or in the interconnected networks connected via WAN (Wide Area Network). In particular, it is an object of this invention to provide a method for detecting an address duplication in a network under control of link state routing.
Accordingly, the present invention is an automatic network-address-duplication detection method for detecting duplicated addresses among a plurality of networked data transmission units under control of link state routing. On one data transmission unit, this automatic network-address-duplication detection method comprises the steps of transmitting a network-wide distribution packet, the transmission comprising the steps of (i) obtaining an address assigned to the one data transmission unit in the network; (ii) reading identification information on the one data transmission unit; adding the obtained address and the identification information to the network-wide distribution packet; and (iv) transmitting the network-wide distribution packet to the network. On another data transmission unit, this automatic network-address-duplication detection method comprises the steps of checking the network-wide distribution packet, the checking step comprising the steps of (i) receiving the network-wide distribution packet;(ii) extracting the identification information and the address of the sending data transmission unit from the received packet; (iii) searching a previously-prepared table for a pair of the identification information and the address of the sending data transmission unit, the table containing a plurality of pairs of the identification information and the address; and(iv) comparing the address contained in the received packet with the address corresponding to other identification information in the table.
The automatic network-address-duplication detection method according to this invention allows an NSAP address duplication to be detected simply by connecting a data transmission unit using the method of this invention to a console terminal for a predetermined period of time. Once the data transmission unit is connected, it sends the LSP containing a unique MAC address to the network at a regular interval and the other data transmission units, upon receiving the LSP, check the received LSP for an address duplication.
Furthermore, the automatic network-address-duplication detection method according to this invention detects duplicated network addresses even if the network contains data transmission units not using the method of this invention. This is because the LSP, which is always distributed to other data transmission units under control of link state routing, contains an MAC address.
The automatic network-address-duplication detection method according to this invention functions even when the network contains a WAN.